Empiricism and Rationalism

Noam Chomsky

Excerpted from Language and Responsibility, Pantheon, 1977

QUESTION: On many occasions you have criticized philosophic and scientific empiricism; Can you state your objections more precisely?

CHOMSKY: In a sense, empiricism has developed a kind of mind-body dualism, of a quite unacceptable type, just at the time when, from another point of view, it rejected such dualism. Within an empiricist framework, one approaches the study of the body as a topic in the natural sciences, concluding that the body is constructed of varied and specialized organs which are extremely complex and genetically determined in their basic character, and that these organs interact in a manner which is also determined by human biology. On the other hand, empiricism insists that the brain is a tabula rasa, empty, unstructured, uniform at least as far as cognitive structure is concerned. I don’t see any reason to believe that; I don’t see any reason to believe that the little finger is a more complex organ than those parts of the human brain involved in the higher mental faculties; on the contrary, it is not unlikely that these are among the most complex structures in the universe. There is no reason to believe that the higher mental faculties are in some manner dissociated from this complexity of organization.

One can say that the dualism introduced by empiricist dogma is methodological rather than substantive. That is to say, it is taken for granted that the body must be studied by the ordinary methods of science but, in the case of the mind, certain preconceptions have been imposed which have virtually removed this study from the domain of scientific inquiry. In fact, this dogmatism seems to be even more striking in the most recent period. Hume, for example, really did his best to show that his elementary principles concerning the acquisition of human knowledge were sufficient to cover an interesting class of cases and challenged his opponents to produce a legitimate “idea” that could not be derived from sense impression by his principles. There is a certain kind of ambiguity in his procedure here, since in part he seems to be engaged in a kind of scientific inquiry, trying to show that certain principles he proposed were in fact adequate to cover the crucial cases, while at other times he relies on these principles to demonstrate that some notion is “illegitimate,” since it cannot be derived by them — an argument that rests on our accepting his not very plausible principles concerning the nature of the mind. Hume regarded the principle of inductive reasoning as a kind of “animal instinct,” which would appear to be an empirical assumption. In modern versions, his assumptions have often been converted into dogma presupposed without serious effort to show them to be valid, or to reply to classical criticisms that were raised against these principles.

There is no reason to believe today that Hume’s principles or anything resembling them are adequate to account for our “ideas” or our knowledge and beliefs, nor to think that they have any particular significance. There is no place for any a priori doctrine concerning the complexity of the brain or its uniformity as far as the higher mental functions are concerned. We must proceed to the investigation of the diverse cognitive structures developed normally by human beings in the course of their maturation and their relation to the physical and social environment, seeking to determine, as best we can, the principles which govern these cognitive structures. Once a certain understanding of the nature of these systems has been obtained, then we can reasonably study the basis on which they are acquired. In my opinion, the little that we know about these questions suggests that the mind, like the body, is in effect a system of organs — we could call them “mental organs” by analogy — that is to say, highly specific systems organized according to a genetic program that determines their function, their structure, the process of their development, in quite a detailed manner; the particular realization of these fundamental principles naturally depends on their interaction with the environment, as in the case of the visual system…. If that is correct, the mind is a complex system of interacting faculties, which do not develop by means of uniform principles of “general intelligence”; it is constituted of “mental organs” just as specialized and differentiated as those of the body.

QUESTION: It is for that reason, doubtless, that you insist on the autonomy of grammar, on the fact that grammatical structures do not depend on other cognitive systems. Does it seem impossible to you to think of language and the structure of knowledge in terms of the same model?

CHOMSKY: I have nothing against comparisons, but I wonder whether we are likely to learn very much by proceeding in this direction. Note that one never tends toward that kind of proposal in physiology; no one suggests that we study the structure of the eye and the heart, and then search for analogies between them. One does not expect to find meaningful analogies. If the mind consists of a system of “mental organs,” in interaction, certainly, but fundamentally different in their structure, we need not expect to find fruitful analogies among them.

To make myself clear, I am not about to propose all this as a new dogma, to replace empiricist doctrine. On the contrary, just as in studying the body, we must simply retain an open mind on this subject. We know a little about a number of cognitive systems, language being the most interesting case at the moment. That small degree of insight seems to me to suggest the preceding conclusions. The important thing, of course, is to determine the deeper principles and the detailed structure of various cognitive systems, their modes of interaction, and the general conditions which each system satisfies. If one finds that these systems are acquired in a uniform manner with very little specific structure, very well. But for the present at least it seems to me that quite different conclusions are indicated. That is what I mean when I say that one need not expect to find analogies.

QUESTION: Nor phenomena of interdependence. However, certain psychologists assert that perception exerts an influence on the potential structure of sentences. An essential aspect of your critique of empiricism is the rationalist hypothesis: the structure of the brain is determined a priori by the genetic code, the brain is programmed to analyze experience and to construct knowledge out of that experience. That may seem shocking …

CHOMSKY: I don’t see anything shocking in that proposition. In physiology, no one has ever accepted anything analogous to empiricist dogma with regard to the mind. No one finds it outlandish to ask the question: What genetic information accounts for the growth of arms instead of wings? Why should it be shocking to raise similar questions with regard to the brain and mental faculties? We are back to the methodological dualism of the empiricists.

QUESTION: That position does not suit contemporary “human sciences.”

CHOMSKY: Especially not behaviorist psychology, or perhaps even Piaget, though his position seems to me obscure in crucial respects. Piaget considers himself to be an anti-empiricist; but some of his writings suggest to me that he is mistaken in this conclusion. Piaget develops a certain “constructive interactionism”: new knowledge is constructed through interaction with the environment. But the fundamental question is evaded: How is this knowledge constructed, and why just this kind of knowledge and not some other? Piaget does not give any intelligible answer, as far as I can make out. The only answer that I can imagine is to suppose an innate genetic structure which determines the process of maturation. Insofar as he considers it wrong to give such an answer, he falls back into something like the empiricism that he wants to reject. What he postulates is nowhere near sufficient, it seems to me, to account for the specific course of cognitive development.

That is not to deny the very great importance of the research that has been conducted by Piaget and his group at Geneva; it has opened up entirely new perspectives in the study of human knowledge. It is primarily the interpretation of their results which seems extremely doubtful to me, in particular their attitude toward what Piaget calls “innéisme,” which seems to me altogether wrong.

In philosophy, the same problems appear in some of the work of Quine, for example. At times, he asserts that theories are developed by induction, which he identifies with conditioning. At other times he says the opposite: theories are not determined solely by conditioning or induction, but involve abstract hypotheses ultimately originating from some innate capacity. In recent years he has oscillated between these two positions.

QUESTION: The tendency of thought which has fought hardest against the independence of grammar as a “mental organ” is without doubt functionalism. It tends to explain the form of language by attributing a determining role to its function. This function is presumed to be communication: everything in language must contribute to communication, to a better communication, and inversely, nothing is linguistic which does not contribute to communication. Isn’t that a fairly accurate portrait?

CHOMSKY: Functionalism holds that the use of language influences its form. This might be understood as a variant of empiricist doctrine about language learning, one that makes very little sense, as far as I can see. But we might understand the fundamental ideas quite differently. For example, George Miller and I suggested about fifteen years ago that there may be a “functional explanation” for the organization of language with grammatical transformations, which would be a well-designed system corresponding to a certain organization of short- and long-term memory, for example.

If one could demonstrate that, it would be interesting. But what does that mean, basically? What would the analogous observation mean for some physical organ, say the heart? To be sure, the heart has a function: to pump blood. One may sensibly say that the structure of the heart is determined by that function. But suppose we ask the ontogenetic question: How does our heart become what it is? How does it grow in the individual from the embryo to its final form in the mature organism? The answer is not functional: the heart does not develop in the individual because it would be useful to carry out a certain function, but rather because the genetic program determines that it will develop as it does.

Every organ has certain functions, but these functions do not determine the ontogenetic development of the organism. Nobody would suggest that a group of cells decides that perhaps it would be a good idea to become a heart because such an organ is necessary to pump blood. If this group of cells becomes a heart, it is due to the information present in the genetic code, which determines the structure of the organism.

There is a place for functional explanation, but it is on the level of evolution. It is possible that a heart develops in the course of evolution in order to satisfy a certain function. Of course, I’m simplifying enormously. But this is a point that is useful to keep in mind: functional explanation does not relate to the way organs develop in the individual.

Let’s go back to linguistics: here comparable remarks can be made. To my knowledge, no functional principle with very great plausibility has yet been proposed. But suppose that someone proposes a principle which says: The form of language is such-and-such because having that form permits a function to be fulfilled — a proposal of this sort would be appropriate at the level of evolution (of the species, or of language), not at the level of acquisition of language by an individual, one would suppose.

QUESTION: As a consequence, insofar as your linguistics is a theory of language and of the acquisition of language by an individual, functionalism cannot be retained as a fundamental principle. Inversely, one might note that the legitimacy of the dependency relation between function and structure is not even a problem for functionalist linguists, because their aim is not to explain the acquisition of language but to describe a linguistic corpus.

CHOMSKY: I doubt that functionalist linguists would accept that characterization. If they mean that ontogenetic development is directed by functional considerations, that seems just as plausible to me as suggesting that the development of the heart in the individual is guided by the utility of having an organ that pumps blood, and about as well supported by the factual evidence. Or, they might say that questions touching on the basis of language acquisition do not concern them. The crucial point, however, seems to me to be that there is no real debate about the validity of functionalism at the generally vague level on which we discuss the hypothetical evolution of the species, or in the study of language change; and there is no sensible way to invoke functional notions as explanatory concepts at the synchronic or ontogenetic level, so far as I can see. It also seems to me important to avoid a certain vulgarization with respect to the use of language. There is no reason to believe … that language “essentially” serves instrumental ends, or that the “essential purpose” of language is “communication,” as is often said, at least if we mean by “communication” something like transmitting information or inducing belief. Someone who claims that this is the essential purpose of language must explain just what he means by it, and why he believes this function, and no other, to be so uniquely significant.

Language is used in many different ways. Language can be used to transmit information, but it also serves many other purposes: to establish relations among people, to express or clarify thought, for play, for creative mental activity, to gain understanding, and so on. In my opinion, there is no reason to accord privileged status to one or the other of these modes. Forced to choose, I would have to say something quite classical and rather empty: language serves essentially for the expression of thought.

I know of no reason to suppose that instrumental ends, or transmission of information about one’s beliefs, or other actions that might reasonably be called “communication” (unless, of course, the term is used quite vacuously), have some unique significance compared with other characteristic uses of language. In fact, what is meant by the assertion that such-and-such is the goal of language, or its essential purpose, is far from clear. … this plurality of modes is characteristic of the most banal and normal use of language.

It is hard to know just what people mean when they say that language is “essentially” an instrument of communication. If you press them a bit and ask them to be more precise, you will often find, for example, that under “communication” they include communication with oneself. Once you admit that, the notion of communication loses all content; the expression of thought becomes a kind of communication. These proposals seem to be either false, or quite empty, depending on the interpretation that is given, even with the best of will. It is all so vague that discussion remains mystifying. I have no idea why such proposals are so often made, frequently with such fervor, or what on earth they are supposed to signify.

The real question is: How does this organism function, and what is its mental and physical structure?

QUESTION: Empiricism (and, in particular, functionalism) has enjoyed an enormous success. In spite of all the demonstrations that have been made of its errors, today it still remains the dominant philosophy. To what do you attribute that success, that power to survive? To a conjunction of ideology and politics?

CHOMSKY: On that point we must be careful, because here we enter into speculation. When certain ideas are dominant, it is very reasonable to ask why. The reason could be that they are plausibly regarded as true, they have been verified, etc. But in the case where they are without empirical foundations, and have little initial plausibility, the question arises more sharply: the answer may actually lie in the domain of ideology. Of course the argument here must be indirect, because we don’t have any direct means of determining the ideological basis for the acceptance gained by a certain doctrine.

Perhaps the instrumentalist conception of language is related to the general belief that human action and its creations, along with the intellectual structure of human beings, are designed for the satisfaction of certain physical needs (food, well-being, security, etc.). Why try to reduce intellectual and artistic achievement to elementary needs? Is the attraction of the several variants of empiricist doctrine based on experimental verification? Hardly. There is no such verification. Does it derive from their explanatory power? No, because they can explain very little. Is it due to some analogy to other systems about which we know more? No. … the systems known to biology are totally different. Animal intelligence seems to be quite different. So too the physical structures of the human organism. The rational hypotheses which we can propose to explain the dominance of empiricist doctrines do not apply.

It should be noted that empiricist doctrine has not merely been “accepted” for a long period, it was hardly even questioned, but rather simply assumed, tacitly, as the framework within which thinking and research must proceed. Perhaps, then, some sociological factor might explain in a natural way why this point of view has been so widely adopted. We can ask ourselves, who accepts and disseminates these doctrines? Essentially, the intelligentsia, including scientists and non-scientists. What is the social role of the intelligentsia? … it has been quite characteristically manipulation and social control in all its varied forms. For example, in those systems called “socialist,” the technical intelligentsia belong to the elite that designs and propagates the ideological system and organizes and controls the society, a fact that has long been noted by the non-Bolshevik left. Walter Kendall, for example, has pointed out that Lenin, in such pamphlets as What Is To Be Done?, conceived of the proletariat as a tabula rasa upon which the “radical” intelligentsia must imprint a socialist consciousness. The metaphor is a good one. For the Bolsheviks, the radical intelligentsia must bring a socialist consciousness to the masses from the outside; as Party members, the intelligentsia must organize and control society in order to bring “socialist structures” into existence.

This set of beliefs corresponds very well to the demands of the technocratic intelligentsia: it offers them a very important social role. And in order to justify such practices, it is very useful to believe that human beings are empty organisms, malleable, controllable, easy to govern, and so on, with no essential need to struggle to find their own way and to determine their own fate. For that empiricism is quite suitable. So from this point of view, it is perhaps no surprise that denial of any “essential human nature” has been so prominent in much of left-wing doctrine.

Analogously, the modern intelligentsia in the capitalist societies — that of the United States, for example — have a certain access to prestige and power by serving the state. So, much the same is true for the liberal intelligentsia in the West. Service to the state includes social manipulation, preservation of capitalist ideology and capitalist institutions, within the framework of state capitalism. In this case as well, the concept of an empty organism is useful. It is plausible that statist ideologues and administrators are attracted by this doctrine because it is so convenient for them, in eliminating any moral barrier to manipulation and control.

These remarks apply only for the last century, more or less. Before that the situation is rather different. Without doubt, at an earlier period, empiricism was associated with progressive social doctrine, in particular, with classical liberalism; although, as we were discussing, that was not always the case. One may recall the ideas of the young Marx, who was far from empiricist doctrine in spirit. Why this link between progressive social thought and empiricist doctrine? Perhaps because empiricism seemed to have — and in a certain way did have — progressive social implications in contrast to reactionary and determinist doctrines, according to which the existing social structures, slavery, autocracy, the feudal hierarchy, the role of women, were founded on unchanging human nature. Against that doctrine, the idea that human nature is a historical product had a progressive content, as it also did, one might argue, throughout the early period of capitalist industrialization.

The determinist doctrines in question maintained that certain people were born to be slaves, by their very nature. Or consider the oppression of woman, which was also founded on such concepts. Or wage labor: willingness to rent oneself through the market is considered one of the fundamental and immutable human properties, in a version of the “human essence” characteristic of the era of capitalism.

In the face of such doctrines as these, it is natural for advocates of social change to adopt the extreme position that “human nature” is a myth, nothing but a product of history. But that position is incorrect. Human nature exists, immutable except for biological changes in the species.

QUESTION: But that is not the same definition of human nature, it is no longer a matter of defining a psychology of individual character.

CHOMSKY: Certainly, we can distinguish between theories that assign a determinate social status to particular individuals or groups by virtue of their alleged intrinsic nature (e.g., some are born to be slaves), and theories that hold that there are certain biological constants characteristic of the species, which may, of course, assume very different forms as the social and material environment varies. There is much to be said about all of these matters. It seems to me that one might suggest, in a very speculative manner, that such factors as the ones I have mentioned entered into the success of empiricism among the intelligentsia. I have discussed this question a bit in Reflections on Language, stressing the crucial and sometimes overlooked point that speculation about these matters of ideology is quite independent of the validity of the specific doctrines in question; it is when doctrines of little merit gain wide and unquestioned credence that such speculations as these become particularly appropriate.

In Reflections, I also mentioned that even at the earliest stages it is not so obvious that empiricism was simply a “progressive” doctrine in terms of its social impact, as is very widely assumed. There has been some interesting work in the past few years, for example, on the philosophical origins of racism, particularly by Harry Bracken, which suggests a much more complex history. It seems that racist doctrine developed in part as a concomitant of the colonial system, for fairly obvious reasons. And it is a fact that some leading empiricist philosophers (Locke, for example) were connected to the colonial system in their professional lives, and that racist attitudes were commonly advanced during this period by major philosophers, among others. It is perhaps not unreasonable to speculate that the success of empiricist beliefs, in some circles at least, might be associated with the fact that they offer a certain possibility for formulating racist doctrine in a way that is difficult to reconcile with traditional dualist concepts concerning “the human essence.” Bracken has suggested, plausibly it seems to me, that racist doctrine raises conceptual difficulties within the framework of dualist beliefs, that is, if they are taken seriously. Cartesian dualism raises what he has called “a modest conceptual barrier” to racist doctrine. The reason for that is simple. Cartesian doctrine characterizes humans as thinking beings: they are metaphysically distinct from non-humans, possessing a thinking substance (res cogitans) which is unitary and invariant — it does not have color, for example. There are no “black minds” or “white minds.” You’re either a machine, or else you’re a human being, just like any other human being in essential constitution. The differences are superficial, insignificant: they have no effect on the invariant human essence.

I think it is not an exaggeration to see in Cartesian doctrine a conceptual barrier — a modest one, as Bracken carefully explains — against racism. On the other hand, the empiricist framework does not offer an analogous characterization of the human essence. A person is a collection of accidental properties, and color is one of them. It is thus somewhat easier to formulate racist beliefs in this framework, although it is not inevitable.

I don’t want to exaggerate the importance of these speculations. But it is worth investigating the question whether colonial ideology did in fact exploit the possibilities made available by empiricist doctrine to formulate more easily the kind of racist beliefs that were employed to justify conquest and oppression. It is unfortunate that the carefully qualified speculations that have been proposed for investigation have evoked a rather hysterical response, and some outright falsification, on the part of a number of philosophers — who are, as Bracken has observed, quite willing to consider, and even advance, very explicit proposals concerning a possible relation between rationalism and various oppressive doctrines, racism among them, thus indicating that it is not the nature of the inquiry but rather its object that they consider intolerable.

I must emphasize again that these speculations, or any others concerning the ideological or social factors that contribute to the success of any doctrine, must be recognized for what they are: speculations which are at best suggestive. Again, questions of this kind arise especially when a doctrine enjoys a great deal of attraction and success among the intelligentsia in spite of little factual support or explanatory value. This is the case with empiricism, in my opinion.

QUESTION: Empiricism thus finds support both from the right and the left … That explains why generative grammar is often attacked by the progressive. intelligentsia, precisely because of your reference to the hypothesis of “innate ideas,” as it is called, that is, the genetic limitations imposed on language. This hypothesis is accused of idealism.

CHOMSKY: That is true, as you say. But the characterization is quite irrational. A consistent materialist would consider it as self-evident that the mind has very important innate structures, physically realized in some manner. Why should it be otherwise? As I have already mentioned, if we assume that human beings belong to the biological world, then we must expect them to resemble the rest of the biological world. Their physical constitution, their organs, and the principles of maturation are genetically determined. There is no reason for supposing the mental world to be an exception. The hypothesis which naturally comes to mind is that these mental systems, unusual in the biological world because of their extraordinary complexity, exhibit the general characteristics of known biological systems. I would emphasize once again that even qualitative considerations of the most evident kind suggest this conclusion: it is difficult to see any other explanation for the fact that extremely complicated and intricate structures are acquired, in a like manner among all individuals, on the basis of very limited and often imperfect data.

QUESTION: Certain psychologists still are trying to make apes talk; as a result, they deny the qualitative difference between human beings and animals, a difference which you have adopted from Cartesianism and restated in the light of modern biology. Do those who are opposed to “innéism” take the same position as these psychologists?

CHOMSKY: I don’t want to speak for others. Let’s consider this question of human uniqueness. Imagine a Martian scientist who studies human beings from the outside, without any prejudice. Suppose that he has a great deal of time at his disposal, say, thousands of years. He notices immediately that there exists on earth a unique organism, whose conditions of life change considerably without corresponding changes in his constitution; that is, modern man. Apes and monkeys live as they lived millions of years ago, while human life changes radically and very rapidly. It is extremely varied, yet there is no corresponding diversity within the human species. Take a child from a Stone Age culture and raise him in New York: he will become a New Yorker. Raise an American baby in New Guinea, and he will become a Papuan “native.” The genetic differences one finds are superficial and trivial, but human beings have the extraordinary characteristic of being able to live in very different ways. Human beings have history, cultural evolution, and cultural differentiation. Any objective scientist must be struck by the qualitative differences between human beings and other organisms, as much as by the difference between insects and vertebrates. If not, he is simply irrational.

Take an even more elementary criterion: proliferation. In that regard human beings are a species with remarkable biological success. Perhaps not if you compare them to insects — or chickens (but here the proliferation in fact results from human intervention) — but compared to higher organisms, monkeys or chimpanzees, for example, they are much more numerous. Thus, in the most elementary respects, human beings are quite different. No scientist could fail to see that.

Even the most superficial observation suffices to show that there are qualitative differences between humans and other complex organisms which must be explained. If our hypothetical Martian observer searches a bit further, he will find that human beings are unique in many respects, one of these being their ability to acquire a rich and varied linguistic system, which can be used freely and in the most subtle and complicated ways, merely by immersion in a linguistic community in which the system is used. It seems to me that a rational observer would conclude that specific qualities of “intelligence,” proper to this species, must be assumed. If he is of an inquiring mind and enterprising, he will seek to determine the genetically fixed mental structures which underlie the unique achievements of this species.

QUESTION: I believe the rejection of “innate ideas” also springs from their association with the Cartesian notion of soul (âme) …

CHOMSKY: That may well be true. But consider this ancient problem of the human soul in its historical context. For Descartes, for example, the existence of the soul is assumed in quite a rational way as a scientific principle. In some respects his argument for the existence of the soul is not very different from Newton’s argument for gravity, as a force of nature. Descartes was wrong, no doubt, but his procedure in itself was not at all unreasonable.

To see this, it is sufficient to pursue the analogy to Newton, though I don’t want to exaggerate its importance. Newton showed that Cartesian mechanics could not account for the movement of heavenly bodies. To explain this movement he postulated a new force: gravity, attraction at a distance; that is, a force which by the criteria of his time was considered to be occult, mystical, because action at a distance violated basic assumptions of mechanics. Newton showed that in this way one could account for the facts, though he too was quite uncomfortable with the “occult force” he was postulating. This postulate became the common sense of following generations, with Laplace and others. An inconceivable idea for pre-Newtonian physics subsequently became part of science because of its remarkable explanatory power.

For his part, Descartes believed — wrongly — that “push-pull” mechanics could explain all phenomena of the natural world, except such things as consciousness and human creativity. Thus to explain what was beyond the scope of his mechanics, he postulated another substance; little else was open to him, given the metaphysics of substance and accident to which he was committed. One can now imagine all sorts of other things, which are not part of his mechanics. But let us suppose that Descartes or the Cartesians could have gone further and invented a mathematics of the mind, a successful explanatory theory. Then their belief would have become part of the science of subsequent generations, like the physics of Newton.

To repeat, the existence of the soul, Descartes’s second substance, is a scientific proposition: it is false, but it is not irrational. Had he elaborated his theory of the soul to an explanatory theory, he might have created a new science to supplement his speculative physiology. He was completely right to propose new principles and to seek out their consequences. One might say that Descartes’s belief that the soul is a simple substance which cannot be analyzed created an obstacle to the development of an explanatory theory of the mind, a theory which in principle might be assimilated to a suitably extended physics — but that is an altogether different question. A convincing rejection of his dualism requires a demonstration that his postulate is useless, or unnecessary because we can explain the properties of the human mind in other ways. Let us then look for such an explanation … It might turn out that we are led to new principles when we inquire into the nature of the mind. It is conceivable, though not demonstrated, that principles entirely different from those of contemporary physics enter into the explanation of mental phenomena. In all these matters one must guard against dogmatism.

QUESTION: To make precise what you are opposing to empiricism, I think it is important to remember that for you the mental organ is that which corresponds to the grammar and not to the language. The structuralists think that one memorizes extended sentences, that is to say, the language (de Saussure’s langue), and that this represents the grammar. But for you, what is constructed in memory as grammar is quite another thing. It is necessary to insist on this difference because so often the set of rules which makes the sentences of a language possible is confused with the language as a set of memorized sequences.

For de Saussure, on the contrary, it was the language — langue — which was deposited in memory. He could not distinguish the memory which we can have of this or that extended sentence from the “memory” — of the grammatical form. The situation is quite different here. The two kinds of memory are different. The construction of the grammar is due to the language faculty. But don’t you think that another confusion can arise because of the ambiguity of the English word language (both langue and language)? Therefore, one could understand that it is the language as langue which is innate …

CHOMSKY: … Which would be absurd, of course; if French were innate, I would speak it …

It is the mechanism of language acquisition that is innate. In a given linguistic community, children with very different experience arrive at comparable grammars, indeed almost identical ones, so far as we know. That is what requires explanation. Even within a very narrow community — take the elite in Paris — the experiences are varied. Each child has a different experience, each child is confronted by different data — but in the end the system is essentially the same. As a consequence we have to suppose that all children share the same internal constraints which characterize narrowly the grammar they are going to construct.

QUESTION: This hypothesis also explains why, when the moment of maturation is passed — adolescence — it is no longer possible to learn a language; the wolf-children never learn to speak, and we speak a foreign language which we have learned late in life with an accent. Without these biological constraints, foreign accent would be inexplicable.

CHOMSKY: Yes, there seems to be a critical age for learning a language, as is true quite generally for the development of the human body. Patterns of growth are determined genetically, for example, sexual maturation, to take a case that occurs long after birth. It would evidently be absurd to maintain that only what one sees at birth is determined genetically.

Even death, to a certain degree, is genetically determined. To say that the genetically determined properties of an organism cannot manifest themselves before the appropriate conditions exist, and that in general the genetic program is spelled out in a way that is partly predetermined and partly influenced by environmental factors, is a virtual truism. In the study of physical development it is a commonplace, and once again, if the methodological dualism of empiricist dogma is abandoned, there is no reason to be surprised by the discovery of similar phenomena in the study of higher mental functions.